Audiogram prediction with machine learning in kids

About Audiogram prediction with machine learning in kids

Hearing evaluation in children is not always complete in one visit because getting usable results depends on the child’s attention and cooperation. As prompt intervention is crucial to limit adverse effects of childhood hearing loss, some uncertainty which remain over audiometric configurations should not delay the hearing aid fitting. Generally, accurate thresholds in at least one low (0.5 or 1 kHz) and one high frequency (2 or 4kHz) can be initially taken with certainty and hearing aid fitting can promptly get started. Therefore, this app was developed in order to predict those missing thresholds in the remaining frequencies using a machine learning algorithm, which is well known in being able to effectively and efficiently predict complex and nonlinear data.